Content Moderation Case Study: Google's Ad Policies Inadvertently Block Religious Organizations From Advertising On YouTube (2019)

Content Moderation

from the unacceptable-content? dept

Summary: Google’s ad service offers purchasers access to millions of users, including those viewing videos on YouTube. But its policies — meant to prevent abuse, fraud, harassment, or targeting of certain demographics — sometimes appear to prevent legitimate organizations from doing something as simple as informing others of their existence.

Chad Robichaux, the founder of Christian veterans support nonprofit Mighty Oaks, wanted to reach out to veterans who might need his services. But his attempt to purchase YouTube ads was rejected by Google’s Ad service for a seemingly strange reason.

According to a screenshot posted by Robichaux to Twitter, Google forbade the use of “Christian” as a keyword. To Robichaux (and many responders to his tweet), this was evidence of Big Tech’s bias against Christians and conservatives.

But the real reason for this block was far less censorial or nefarious, if no more explicable. According to YouTube (which reached out directly to Robicheaux), the aim isn’t to keep Christians from advertising, but rather to prevent advertisers from targeting users on the basis of their religion. Unfortunately, Google’s policy doesn’t exactly make that clear, instead stating that ads cannot contain “religious basis” content if the purchaser is engaging in personalized advertising.

Decisions to be made by Google:

• Does blocking certain keywords make some ads impossible to place, no matter what audience is targeted or where the content may appear?

• Is it ok for advertisers to target these groups if the users have already self-identified as being members of these groups? Would it be ok if users could explicitly opt in to being targeted in this way?

• Is clarification or simplification of the rules needed to ensure accidental blocking or further misunderstandings are avoided?

• Should advertisers be given more guidance on how to craft ads/seek users to prevent violations?

Questions and policy implications to consider:

• Does having control of a majority of the advertising market lower the quality of assistance users receive from Google given the limited options available to them elsewhere?

• Does increasing the number of keyword restrictions result in fewer successful ad placements and lower ad sales?

• Does “protecting” users from personalized ads using certain keywords result in users see more irrelevant ads?

Resolution: The confusion was (somewhat) cleared up by YouTube’s direct contact with the concerned ad buyer. But other confusion still remains since the policies guiding ad purchasing/ad construction are far from straightforward. Allegations of bias were off-base. Instead, it was simply Google enforcing its policies, which would have made it equally impossible to use any other religion as a keyword.

Originally posted on the Trust & Safety Foundation website.

Filed Under: ads, content moderation, religious targeting, youtube
Companies: google, youtube